Successful In Situ Targeting of Pancreatic Tumors in a Novel Orthotopic Porcine Model Using Histotripsy

Objective: New therapeutic strategies and paradigms are direly needed to treat pancreatic cancer. The absence of a suitable pre-clinical animal model of pancreatic cancer is a major limitation to biomedical device and therapeutic development. Traditionally, pigs have proven to be ideal models, especially in the context of designing human-sized instruments, perfecting surgical techniques and optimizing clinical procedures for use in humans. However, pig studies have typically focused on healthy tissue assessments and are limited to general safety evaluations because of the inability to effectively model human tumors. Methods: Here, we establish an orthotopic porcine model of human pancreatic cancer using RAG2/IL2RG double-knockout immunocompromised pigs and treat the tumors ex vivo and in vivo with histotripsy. Results: Using these animals, we describe the successful engraftment of Panc-1 human pancreatic cancer cell line tumors and characterize their development. To illustrate the utility of these animals for therapeutic development, we determine for the first time, the successful targeting of in situ pancreatic tumors using histotripsy. Treatment with histotripsy resulted in partial ablation in vivo and reduction in collagen content in both in vivo tumor in pig pancreas and ex vivo patient tumor. Conclusion: This study presents a first step toward establishing histotripsy as a non-invasive treatment method for pancreatic cancer and exposes some of the challenges of ultrasound guidance for histotripsy ablation in the pancreas. Simultaneously, we introduce a highly robust model of pancreatic cancer in a large mammal model that could be used to evaluate a variety biomedical devices and therapeutic strategies.


Introduction
Histotripsy has recently been suggested as an alternative tumor ablation modality that could prove useful in treating pancreatic cancer [1].Histotripsy is a non-ionizing, non-thermal and image-guided non-invasive tissue ablation technique capable of treating solid tumors [2].Histotripsy using high pressure (>10 MPa) and very short duration (<20 μs) generates a distinctive cavitation "bubble cloud" at the focus of which expansion and collapse create high stress to completely disintegrate target tissue into an acellular homogenate with no residual cellular structures [3−8].As a non-thermal modality, histotripsy has been successfully used to ablate tumors and healthy tissues near vital organs, such as the spleen, liver, major vessels, nerves, and bile ducts without causing damage [1,4,9−12].Histotripsy has also been used in the highly vascular liver to generate full ablation successfully and consistently [13,14].This includes the completion of the first-in-human use of histotripsy to treat hepatic tumors, reported earlier in 2022 [15].This complements several pre-clinical rodent studies and veterinary clinical trials that have reported successful ablation of osteosarcoma, cholangiocarcinoma, pancreatic tumors and spontaneous soft tissue [1,9,10,16−19].
For all of the tumor ablation modalities, including histotripsy, key data are required for the establishment of ablation parameters for each targeted tissue.Ex vivo tumor tissue and healthy organs typically lack accurate mechanical properties and postmortem tissues usually become unsuitable for ablation modality testing [20,21].Murine models have certain advantages that make them appealing, such as ease of housing and handling, cost, genetic similarities to humans and availability of numerous established lines of interest [22−24].However, the significant discrepancy in size, anatomy and genetic makeup between mice and humans and ablation volume are major challenges when translating ablation findings from mice to humans.Veterinary clinical trials can also be conducted to evaluate treatment effectiveness on spontaneous tumors.However, these studies can often lack statistical power, the ability to conduct robust and highly controlled experimental studies can be limited and reagent availability for mechanistic studies can be a limitation.
Pigs have proven to be a useful model for biomedical device development and tumor ablation trials, especially in the context of designing human-sized instruments, improving surgical techniques and optimizing clinical procedures for use in humans [4,9,20,25−27].However, to date, pig studies have focused on healthy tissue assessments and are limited to general safety evaluations.Although critical for translation to human patients, the focus on healthy tissues is a significant limitation.As mentioned above, healthy tissue studies do not account for physical, mechanical, physiological and biological differences that exist between healthy tissue and tumor.Thus, to circumvent these limitations, we generated unique RAG2/ IL2RG-deficient immunocompromised pigs [28] and previously determined their ability to be engrafted subcutaneously with human xenograft tumor cell lines [20].Here, we extend these prior subcutaneous studies and conduct orthotopic engraftment of human Panc-1 cells in the pig pancreas.We demonstrate the utility of this model in accurately and consistently generating tumors in specific locations of the pancreas and further characterize the tumors generated with those derived from comparable mouse pancreatic tumors (Pan02) and specimens from human patients.To further illustrate the utility of this model in ablation therapy development, we successfully targeted a subset of the pancreatic tumors with histotripsy and determined for the first time the feasibility of using histotripsy to treat pancreatic cancer.This work serves as an effective proof-of-concept study and has identified both strengths and challenges in targeting pancreatic tumors.

Generation of immunocompromised pigs by RAG2/IL2RG deletion
RAG2/IL2RG double-knockout pigs were generated using CRISPR/Cas9 as previously described (Figs.S1 and S2, online only) [28−30].A sterile hysterectomy method was used to deliver piglets from the sow that were immediately aseptically transferred to isolators for housing under germ-free conditions [31].All study methods were in compliance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals and with the Virginia Tech Institutional Animal Care and Use Committee (IACUC).The immunocompromised status of each pig was validated through genotyping and sequencing to confirm RAG2/IL2RG double-knockout status (Fig. S3, online only).

Ethics
All human specimens were collected and used according to institutional review board (IRB) guidelines following institutional approval.All animal experiments were approved and carried out in accordance with the Virginia Tech Institutional Animal Care and Use Committee under IACUC Protocol 19-196-CVM, approved December 15, 2021, and22-047-CVM, approved March 31, 2022.

Surgery and implantation of Panc-1 human pancreatic cancer cells into pig pancreas
Human pancreatic ductal epithelial carcinoma Panc-1 cells (ATCC, Manassas, VA, USA) were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum and 1% Normocin.To generate a tumor with a defined margin, cells were resuspended in Matrigel at a concentration of 6 × 10 6 /100 μL and kept on ice.One-week-old pigs were individually placed under anesthesia for surgery and injected in the pancreas with 100 μL of Panc-1 cells; tumors were permitted to grow for about a month (25−30 d) before histotripsy (Figs.S1, S2B and S2C; Table 1).In the present study, the pigs were treated as follows: 1 pig was used as a control (Table 1, no surgery or tumor injection); 6 pigs were used to validate the surgery, characterize orthotopic tumor progression and identify optimal acoustic windows for ultrasound visualization and histotripsy treatment (Table 1); 3 pigs were used to conduct proof-of-concept studies using histotripsy (Table 1).

Ex vivo histotripsy treatment
Histotripsy was tested in the excised Panc-1 tumors from the RAG2/IL2RG pigs ex vivo using the same transducer.Tumors were de-gassed, embedded in gelatin phantoms and treated as previously described [6,20].Histotripsy was applied to the first and second tumor samples at ~14 MPa peak negative pressure with an ~1.5 cm diameter spherical ablation volume.In addition to the excised tumor samples from pigs treated with histotripsy, a human excised pancreatic tumor sample (stage IIA, freshly received in 0.9% de-gassed saline) was also treated and compared with the Panc-1 tumors.

In vivo histotripsy treatment
One week prior to treatment, pigs were fed a diet that included sweetened custard with simethicone and bisacodyl to minimize the intestinal contents and gas [32,33].All animals were sedated using 2−4 mL/kg of Telazol−Ketamine−Xylazine (TKX) and kept under anesthesia throughout all procedures.The pigs' weight range at the time of treatment was 6−7 kg.Histotripsy was performed by targeting each orthotopic Panc-1 tumor in vivo/in situ using a custom 32-element 500 kHz therapy transducer driven by a custom high-voltage field programmable gate array (FPGA) board (Altera DE0-Nano Terasic Technology) and pulser to deliver short histotripsy pulses <2 cycles.The therapy transducer has a focal depth of 78 mm with transverse and elevational aperture dimensions of 128 and 112 mm, respectively.Further device specifications are previously reported [1].A robotic micropositioner connected the transducer to a histotripsy treatment cart (HistoSonics) with a 3 MHz curvilinear imaging probe (Model C52, Analogic Corp., Peabody, MA, USA) coaxially aligned with the therapy transducer to allow for real-time imaging feedback for treatment monitoring as detailed previously [1].To target the pancreatic tumors, a veterinary radiologist (M.E.) used free-hand ultrasound imaging (Model C5-2, Analogic Corp.) to identify the tumors and mark the best acoustic window for histotripsy on the abdomen of the pig (Fig. S2D).Single-cycle histotripsy pulses at a pulse repetition frequency (PRF) of 500 Hz were applied through an automated volumetric spherical ablation strategy that moved through the tumor volume, twice delivering a total of ~1000 pulses per treatment point (PPP).To allow overlap of the bubble cloud at each treatment point, points were spaced by 3.5 mm in the axial direction and 1.5 mm in the lateral and elevational directions.Pressure was slowly increased until a bubble cloud was generated in the targeted tumor.The in situ focal pressures could not be directly measured non-invasively.With 3−4 cm of overlying tissue and using 0.5 dB/cm-MHz attenuation for the overlying tissue, the in situ peak negative pressure was estimated to be ~31.2MPa.However, it is worth noting that this is likely an overestimate of the actual in situ pressure as this estimate does not include pressure loss caused by gas blockage and tissue aberration, which recent work has indicated can be significant for intra-abdominal histotripsy procedures [34].Tumors were treated twice with single-cycle histotripsy pulses at a PRF of 500 Hz.A full necropsy was performed immediately after the treatment by a veterinary pathologist (S.C.O./K.E.).

Histopathology
All tissues were fixed in formalin and stained with hematoxylin and eosin (H&E) or trichrome.Histology sections were prepared at 5 mm thickness with 20 or 50 mm intervals from multiple sections of tumor to find start, middle and end of the ablation zone based on the assessment of a board-certified pathologist.To verify the tissue ablation, regions of cellular damage within and outside the ablation zone were assessed.Trichrome-stained images were scored using Fiji.Images were subjected to color deconvolution into trichrome using Fiji, and only blue-stained collagen sections were analyzed, as previously described [35].

Blood chemistry
Blood was collected from animals to evaluate pancreatitis.Blood samples were collected from the jugular vein using un-supplemented BD Vacutainer blood collection tubes, and serum was separated following centrifugation to remove the red blood cells.Levels of amylase and lipase were evaluated in the serum.

Successful generation of a pig orthotopic pancreatic tumor model using the Panc-1 human pancreatic cancer cell line
We generated 10 RAG2/IL2RG pigs using the CRISPR/Cas9 system as previously described [9,28,29].Briefly, we used 1265 in vitro fertilized oocytes where the porcine RAG2 and IL2RG genes were targeted using CRISPR/Cas9 prior to in vitro culture.Four days after in vitro fertilization (IVF), 179 developing embryos between the eight-cell and morula stages were transferred to surrogate gilts.Genomic DNA collected from ear notches revealed that RAG2 and IL2RG were successfully targeted in all 10 piglets (Fig. S3).Our engraftment rate for the orthotopic Panc-1 cell line in the RAG2/IL2RG immunocompromised pigs is 90.9%.Metastatic tumors were noted in one animal (Table 1; Fig. S4).Of note, none of the pigs developed pancreatitis, based on normal blood ranges of amylase and lipase (Fig. S5, online only).Our results illustrate our ability to successfully engraft orthotopic human Panc-1 tumors in RAG2/IL2RG knockout pigs.

Ex vivo excised tissue used to define histotripsy parameters for pancreatic tumor ablation
To refine and optimize the histotripsy parameters for in vivo studies, orthotopic tumors were excised from pigs and human patients .Tumors excised from mice were ~1 cm in diameter (not shown), orthotopic pig tumors were ~3 cm in diameter and the human tumor was ~3 cm in diameter (Figs.1A and 2A).The excised pig tumors were pathologically characterized as corresponding to stage IIA human pancreas ductal adenocarcinomas.The histotripsy parameters were gradually increased until a bubble cloud was identified in the targeted region of the tumor on ultrasound imaging.This was noted at ~13.6−14.1 MPa peak negative pressure, which was applied to an estimated 1, 1.4 and 1.5 cm diameter spherical ablation volumes in the human and two excised pig tumors, respectively (Figs. 1C, 1D, 2C and 2D).Histopathology revealed that tumor tissue was successfully ablated in tumors from both pig and human patients under these conditions (Figs. 1E, 1F and 2E-G).Ablations were confirmed to be partial, with the remaining Panc-1 cells generally dispersed in small pockets throughout the ablation zone.Ablated tissues were characterized by loss of recognizable tissue architecture and replacement by fine basophilic (bluish) debris.Larger pockets of cells were visible along the margins, with a clear demarcation between the ablation zone and untreated tumor (Figs.1E, 1F and 2E-G).This is consistent with previously reported findings from the use of histotripsy to treat ex vivo mouse Pan02 subcutaneous tumors with partial ablation [18].

Histotripsy successfully ablates orthotopic pancreatic tumors in situ
For all pigs, prior to treatment, pancreatic tumors were visualized using ultrasound imaging using a freehand ultrasound imaging probe with pressure applied to the abdomen to displace bowel gas (Fig. 3A).Although some tumors were readily identifiable by ultrasound and effectively targeted with histotripsy, some of the pancreatic tumors could not be consistently visualized by the co-axially aligned imaging probe because of the bowel gas and the inability to apply significant pressure.Under these conditions, the focus of the histotripsy transducer was aligned to pre-marked anatomical landmarks.Histotripsy was applied at approximately >31.2 MPa peak negative pressure.These in vivo pressure levels were notably higher than the ex vivo histotripsy ablations, which used 14 MPa peak negative pressure.The difference in pressures can likely be attributed to aberration and gassy effects in the overlying tissues exhibited in vivo resulting in the higher estimated pressures reported for generating the bubble clouds in vivo.
Histotripsy was applied through an automated volumetric ablation strategy that moved through the tumor volume twice with a total PPP for the complete treatment of ~1000.The first and second pigs were treated with a 1.75 and 1.90 cm diameter spherical ablation volumes, respectively.With tumors ~2 cm in diameter in each pig, these treatments corresponded to ~67% and ~86% of the tumor volumes targeted with histotripsy, respectively.During treatment, histotripsy cavitation was audible, but the bubble cloud was not clearly identifiable on ultrasound imaging in all tumors.
After treatment and necropsy, tumor characteristics were noted, and both targeted and non-targeted ablation zones were evaluated (Fig. 3B-D).Gross morphology of the treated pancreatic tumors revealed a well-defined histotripsy ablation zone, which represented a partial tumor ablation (localized hemorrhage) (Fig. 3D).Ablation zones were approximately 1.0 cm in size (Fig. 3D).No off-target damage to surrounding organs was noted.A successful but partial ablation in the pancreatic tumor was verified by histopathology evaluation (Fig. 4A, 4B).Despite issues with gas in the gastrointestinal tract and the lack of robust image guidance, these studies indicate for the first time that histotripsy, even under suboptimal conditions, can ablate pancreatic tumors in vivo/in situ under clinically relevant conditions.
Orthotopic Panc-1 tumors in pigs share similar pathological characteristics with other common cell line−based pancreatic tumors before and after histotripsy treatment The majority of studies evaluating therapeutic approaches targeting pancreatic cancer use rodent models, such as the mouse Pan02 model [18,36−38].Thus, we next sought to compare the human Panc-1 tumors grown in the immunocompromised pigs with the murine Pan02 tumors propagated in wild-type mice [9,18,20].In general, the Panc-1 morphological characteristics of tumors from both immunocompromised pigs and Pan02 tumors from immunocompetent mice were highly similar (Fig. 4A-D).Both tumors were both densely cellular, although we observed areas of central necrosis of irregular size within the Pan02 murine tumors that were not observed in the Panc-1 tumors (Fig. 4C).In both tumor types, spindle-shaped cells were irregularly arranged without any distinct pattern.For histotripsy treatment, we attempted to use a partial treatment parameter that was not designed to treat the whole tumor (Fig. 4B, 4D).Here, we report that subcutaneous murine Pan02 tumor targeting and ablation (Fig. 4D) have improved efficacy compared with the orthotopic Panc-1 tumor (Fig. 4B) based on the generally welldefined and larger ablation zones in treated murine tumors.Previous studies have established the micro-heterogeneity of collagen in human pancreatic tumor types with the collagen density associated with the tumor biomechanical stiffness and inversely related to vascular perfusion [39,40].The Panc-1 tumors in the pig pancreas were stiffer than the normal pancreas measured with B-mode ultrasound (Fig. 3A) and were readily identifiable grossly from the surrounding pancreas tissue during necropsy.B-mode (brightness mode) images reveal the relative stiffness of lesions compared with the stiffness of surrounding softer tissue.Stiffer areas deform less easily than their surroundings and are depicted in darker (hypo-echoic) images, whereas softer areas deform more easily than their surroundings and are depicted as lighter areas in the images.Malignant masses typically appear dark and have high contrast against the background in pancreas.The hypoechoic region is clearly visible in Figure 3A (yellow arrow, within stars) and the masses were readily identified by touch during the necropsy.The healthy pig pancreas is clearly divided into multiple lobules by connective tissue septae (Fig. 5A).Tumor tissue was characterized by a loss of normal lobular architecture and replacement by tumor cells embedded in an abundant collagenous stroma (blue) (Fig. 5B).After histotripsy treatment, significantly less collagen was observed (B vs. C in Fig. 5F graph).Of note, the untreated pancreatic tumor from the human patient had significantly higher collagen than the untreated tumor grown in pig pancreas (Fig. 5B, 5D, 5F).However, ex vivo histotripsy treatment of the tumor tissue from the human patient exhibited reduced collagen when compared with before treatment (Fig. 5B, 5D, 5F).Taken together, our data suggest that the orthotopic pig model depicts nodular tumor formation like the mouse model and better reflects the collagenous stroma formation seen in the human patient tumor.Data also suggest that histotripsy is able to effectively ablate tumor tissue, along with collagenous materials and stroma.

Discussion
Histotripsy is a promising ablation modality because of its non-thermal, non-ionizing and non-invasive nature.Nevertheless, along with other highly promising ablation modalities, clinical translation of histotripsy to treat pancreatic cancer has been hindered by the lack of vigorous and clinically relevant pre-clinical animal models.In this study, we establish an orthotopic and highly reproducible human pancreatic tumor model in swine that accurately and effectively mimics many of the clinical challenges expected to be faced in human patients.The ability to surgically engraft tumors within the pancreas in specific areas that are difficult to treat and/or in areas adjacent to critical anatomical structures in the pancreas allows for highly robust limitation, proof-of-concept and proof-of-principle testing.
Using this unique model, we report for the first time the successful histotripsy ablation of pancreatic tumors in situ under clinically relevant conditions.Histotripsy successfully destroyed tumor tissue in the ablation zone with no adverse effects observed during the monitoring period.Within the ablation zone, individual and small pockets of Panc-1 cells were still observed, mostly at the margins of the treatment zones and in areas outside of the targeted region of the organ.Gas in the overlaying gastrointestinal tract was a significant limitation.The lack of a clear bubble cloud in some of the targeted tumors, in part because of gas, may necessitate more invasive approaches to remove gas in difficult-to-visualize targets.
Although we were successful in creating ablations and significantly debulking the tumors, we did not achieve a full ablation in any of the targets.Similar to human patients, the Panc-1 tumors generated in this study exhibited a dense stromal region that has altered the mechanical properties of the tumors and appeared to reduce histotripsy effectiveness.This is typical of pancreatic tumors, which can be highly heterogenous in terms of anatomy, cellular constituents and stromal density [41,42].Thus, each of these characteristics should be taken into account for not only histotripsy, but any local tumor ablation modality.These data will be used to inform, optimize and refine our approaches using these unique animal models in future studies.
There is currently extensive debate in the tumor ablation field regarding the benefits of partial ablation versus full ablation and the order of operations in terms of treatment timing and surgical intervention [43,44].Much of this stems from the desire to optimize the host immune response following treatment.Local tumor ablation can augment immune system activation to promote the clearance of the remaining residual cancer cells in the treatment zone and at the margins following ablation [45−52].Local tumor ablation can also promote the systemic anti-tumor adaptive immune response to target the micro-metastatic lesions in the primary organ and the metastatic tumors at distal sites [45−50].Thus, the benefits of ablation modalities such as histotripsy have exciting potential to extend beyond the local ablation zone, but more work is necessary to better define the underlying biology of immune system activation to inform the ablation modality treatment parameters.However, what is clear from the data presented here is that even partial histotripsy ablation is capable of significantly reducing the stromal niche, specifically the collagen composition, of the Panc-1 tumors (Fig. 5F).This suggests that histotripsy may be effective in increasing therapeutic access to the targeted tumor.The dense stroma is a significant limitation for many of the current frontline therapeutics for pancreatic cancer [53−55].Thus, by degrading this stromal compartment, histotripsy may be an effective co-therapeutic approach for many of these current and emerging therapeutic approaches that previously had limited effectiveness.
The use of local tumor ablation modalities, such as histotripsy, to induce a robust and reproducible systemic, anti-tumor immune response to effectively eliminate cancer cells in the untreated margins and at metastatic sites is an emerging area of research interest [18].Unfortunately, while the lack of an intact immune system enables human tumor cell line and PDX engraftment in the RAG2/IL2RG pigs described here, this model does not effectively allow assessments of the immune system niche.Thus, as we report here, although this model is ideal for reproducing mechanical and physical tumor properties, it is limited in terms of evaluating the use of co-therapeutic approaches based on immune system augmentation.For example, prior work has had promising results in mouse models combining histotripsy approaches with immune checkpoint inhibitors [56,57].These studies reveal improved immune checkpoint blockade therapy of poorly immunogenic pancreatic tumors.To better evaluate the pancreatic cancer immune niche in pigs, recent work has resulted in the development of a porcine pancreatic cell line that can be used in immunocompetent animals [58,59].However, the model still requires optimization before it can be fully deployed [58,59].Limitations of this immunocompetent model currently include a lack of investigation of the tumor microenvironment and stroma formation compared with human pancreatic tumors [58,59].To circumvent these issues, the use of transgenic pigs with Cre-inducible TP53 R167H and KRAS G12D mutations has also been developed to model pancreatic cancer in immunocompetent animals [60].The tumors generated in these "Oncopigs" are undifferentiated carcinomas with a significant inflammatory component similar to that of pacreatobiliary carcinomas in human patients [60].However, there are several limitations to the Oncopig model [58,60,61]: surgery is necessary to place a gelatin sponge necessary for the delivery of the adenoviral vector for Cre induction, the tumors themselves appear highly inflammatory as opposed to the generally immuno-suppressive tumor microenvironment commonly observed in human patients, the tumors are poorly differentiated and hypovascular, cost and scale are common concerns with this model, the reliance on the adenovirus vector can create tumors that are difficult to predict and localize and tumors can take over a year to reach <1 cm in size [58,60,61].It should be noted that in our current study, the size of the pig could be considered a limitation.Here, we use piglets rather than adult pigs.The immunocompromised status of the piglets allows us to grow human PDAC tumors; however, their immunocompromised status requires that they be housed in isolators that place limits on animal size.The animals used in the current studies were allowed to grow to the maximum size possible for the isolators used.Based on the successful results of the current study, future studies will transfer animals to larger isolators to allow for the animals to reach more clinically relevant sizes.Thus, although the pancreatic cancer pig model characterized in the current work has limitations, this model also has several strengths over the previously described models that should be considered for studies moving forward.

Figure 1 .
Figure 1.Histotripsy optimization using ex vivo tumor treatment.(A) Gross image of the excised Panc-1 tumor.(B) Excised Panc-1 tumor inside Matrigel block.(C, D) Ultrasound image of the Panc-1 tumor inside the block (C) before and (D) after treatment (red arrow points to hypo-echoic region after treatment).(E, F) Histopathology assessments were used to characterize the ablation zone after hematoxylin and eosin staining using images at (E) 1.1 ×, bar = 1000 μm and (F) 10 ×, bar = 100 μm.Ablated cells stain more pink in hematoxylin and eosin than healthy cells.

Figure 2 .
Figure 2. Human pancreatic tumors have an ablation profile similar to that of the pig-derived Panc-1 tumor.(A) Gross image of the excised human pancreatic tumor.(B) Excised tumor embedded in Matrigel block.(C, D) Ultrasound images of the pancreatic tumor inside the block (C) before and (D) after treatment (yellow arrow points to hypo-echoic region after treatment).(E−G) Histopathology assessments were used to characterize the ablation zone after hematoxylin and eosin staining using images at (E) 4 ×, (F) 20 × and (F) 40 ×, with scale bars of 500, 50 and 25 μm.respectively.Insets focus on the ablation zone.

Figure 3 .
Figure 3. Histotripsy successfully ablated orthotopic Panc-1 tumors in the pig pancreas under clinically relevant conditions.(A) Orthotopic Panc-1 tumors were readily identifiable using ultrasound and were engrafted in areas adjacent to critical structures (such as the splenic vein, white arrow) or in areas visually devoid of critical structures to allow for limitation testing.Tumors are marked with the white "+" and yellow arrows which are more hypoechoic region in B-mode ultrasound.(B, C) Tumors were verified and characterized after necropsy, both (B) in the pancreas (yellow dashed line) and (C) after excision.(D) Histotripsy ablation zones were readily identified in gross specimens after treatment, with the tumor outlined by the yellow dashed line and yellow arrow and the ablation zone

Figure 4 .
Figure 4. Histopathology characterization of treated and untreated Panc-1 tumors with mouse Pan02 tumors.(A, B) Human Panc-1 tumors after either (A) no treatment or (B) histotripsy treatment and evaluation by histopathology using hematoxylin and eosin−stained sections.Histotripsy treatment resulted in dispersed ablation zones within tumor.(C) Panc-1 tumors exhibited significantly different histopathological features compared with subcutaneously engrafted mouse Pan02 tumors engrafted in immunocompetent C57Bl//6 mice.Subcutaneously engrafted mouse Pan02 tumors exhibit central necrosis (indicated by yellow arrow) not found in human Panc-1 tumor grown in pig pancreas.(D) Histotripsy

Figure 5 .
Figure 5. Histotripsy significantly reduced the stromal compartment and collagen levels in the orthotopic Panc-1 tumors in the pig pancreas in situ and also ex vivo treated human patient tumor.(A−C) Healthy and tumor specimens were stained with trichrome for histopathology assessments.(A) Trichrome assessments of the healthy pig pancreas revealed very minimal collagen content (arrows indicate lobules of healthy pancreas).(B) The orthotopic Panc-1 tumors exhibited dense collagen staining (blue).(C) Histotripsy significantly reduced the collagen levels and generated large tracks of collagen and stroma-free areas in the tumor tissue.(D) Patient tumor exhibits very high content of collagen.(E) Ex vivo partially treated patient tumor exhibits reduced collagen content in the treated areas.From left to right, magnifications are 4 ×, 20 × and 40 ×, and bars are 500, 50 and 25 μm, respectively.(F) Fifteen images from groups A−E were scored for blue (collagen) staining using FIJI.Scoring indicates reduced collagen after histotripsy treatment (bar B vs. bar C and bar D vs. bar E). ****p < 0.00001, ***p < 0.0001.